Sequence & measure
Convert biological material into digital observations while tracking read quality, coverage, experimental design, and measurement uncertainty.
Biology · algorithms · statistics · databases
Bioinformatics
Bioinformatics builds computational representations of biological data so sequences, structures, expression patterns, populations, and experiments can be compared at scales impossible to inspect by hand. Its job is not merely to compute, but to preserve the biological meaning and uncertainty of what was measured.
The computational problem
A genome is data, but biological meaning is not written between the letters for us.
Sequencing can produce millions or billions of symbols. Bioinformatics asks which observations correspond, where features begin and end, what patterns are unusual, and which claims remain plausible after uncertainty and bias are considered.
Sequence alignment workstation
An alignment proposes which symbols correspond.
Switch examples, then click an aligned column. The goal is not to memorize a score. It is to distinguish a match, a substitution, and a gap before asking what the difference might mean biologically.
reference
relation
sample
Column 1
AAmatch
A substitution places different symbols in corresponding aligned positions. Whether a nucleotide change matters biologically depends on where it occurs and what function that region has.
matches
10
substitutions
2
gap columns
0
ungapped identity
83%
Important: an alignment is a model, not ground truth. Different scoring rules, reference choices, repetitive regions, and evolutionary assumptions can produce different plausible alignments.
Sequence to inference
01
02
Compare & search
Align sequences, search databases, assemble fragments, and identify patterns that may represent shared ancestry, conserved function, or technical artifacts.
03
Annotate & represent
Attach coordinates, features, genes, variants, structures, pathways, and relationships to otherwise anonymous strings and records.
04
Model & test
Use statistical and computational models to separate signal from noise, estimate uncertainty, compare hypotheses, and predict biological behavior.
05
Infer biology
Translate computational results back into biological claims about function, evolution, disease, regulation, ecology, or experimental mechanism.
Sequence ≠ function
A DNA or protein sequence can constrain hypotheses about function, but sequence similarity alone does not prove that two molecules behave identically.
Database ≠ nature
Reference databases reflect what has been sampled, curated, named, sequenced, and deposited. Missing or biased data can shape the conclusions.
Prediction ≠ validation
A computational prediction becomes stronger when independent experimental, clinical, ecological, or structural evidence supports it.